OpenAI still has a WebRTC Problem - Media over QUIC
Not investment advice. Muhabbit scores how much a story may matter, not which way to trade.
OpenAI is still relying on WebRTC for its voice AI products, a protocol critics argue is ill-suited to machine-speed audio because it drops late packets and disables retransmissions by default for audio, causing voice agents to lose or mishear user speech during packet loss. The critique, from streaming engineer kixelated, follows outreach from OpenAI WebRTC experts Sean DuBois and Justin Uberti after his earlier viral rant on the topic. The core issue is that WebRTC was built for human conference calls, where latency above roughly 30ms is unacceptable, so it favors real-time delivery over reliability—a tradeoff that wastes tokens and forces users to repeat themselves when connections degrade. The article points to Media over QUIC as the emerging alternative designed for silicon-clock-rate audio consumption, and argues the problem is fixable rather than inevitable. What to watch is whether OpenAI and other frontier labs migrate their voice stacks off WebRTC and how quickly Media over QUIC matures as a replacement.
Muhabbit's AI wrote this summary in its own words from Hacker News's report. It can get details wrong, so check the source.
Named in this story
WebRTC, OpenAI, Media over QUIC, Sean DuBois, Justin Uberti